Recently, immunotherapy has been used as a safe therapy for the treatment of cancer, rheumatoid, etc. In the immunotherapy of cancer, an antibody showing a cytotoxic activity specifically upon cancer cells is used. While it is recognized that an antibody drug comprising such antibody will show high and safe therapeutic effects with little side effects, it has a problem that it would cost much since said drug needs to be produced by using established animal cells.
As a result, it has been a worldwide trend to produce a low molecular-weight antibody such as a single-chain antibody (scFV) that contains VH and VL of a certain antibody in a single-chain polypeptide. Such low molecular-weight antibody can be economically produced by E. coli. However, it is concerned that its half life in a body will be decreased due to its low molecular weight, reducing the period of effecting medical benefits. Also, it is a problem that affinity of such low molecular-weight antibody with monovalence is lower than that of a full antibody such as IgG with polyvalence for a target antigen. Furthermore, as a main mechanism of an action of the antibody drug is considered to be an antibody dependent cytotoxic activity (ADCC) via Fc region, it is concerned that the ADCC of the scFv that has no Fc region would be low. Non-Patent Document 1 may be referred to with respect to the scFv.
Accordingly, a bispecific antibody with a low molecular weight has been developed, which can cross-link between cancer cells and immune cells. Only one of such bispecific antibody with a low molecular weight, called “BiTE”, which consists of two fragments of scFv linked with each other in tandem, has been now brought into a clinical trial (Science 2008 Aug. 15:321 (5891): 974-7). However, as the BiTE is produced by using animal cells, its production cost and yield have become problematic. Furthermore, it was reported that it was difficult to prepare the tandem scFv-type bispecific antibody with a low molecular weight such as BiTE from soluble fraction of E. coli (J Mol Biol, 2003 330(1):99-111).
Among antibodies with multiple specificities, an antibody with bispecificity (Bispecific Antibody: BsAb) has been studied intensively. The bispecific antibody can bind specifically to two different kinds of antigens so that it will be utilized as a therapeutic agent having a specific anti-cancer effect. A diabody (Db) is a minimum unit of the above bispecific antibody. It was developed by utilizing the property that the variable region in a heavy chain (VH) and the variable region in a light chain (VL) derived from the same parent antibody will form a hetero-dimer through non-covalent bond (Non-Patent Document 2).
The diabody-type bispecific antibody is characterized by having low immunogenicity and high infiltrating activity into tumor tissues due to its low molecular weight (ca. 60,000), and by being able to be easily mass-produced at a low cost with use of microorganisms such as E. coli, and to be easily altered in function by means of genetic engineering.
The present inventors already found that the diabody-type bispecific antibody (Ex3) that was produced by utilizing an anti-human EGF receptor 1 (Her 1) antibody 528 and an anti-CD3 antibody OKT3, and its humanized diabody-type bispecific antibody (referred to as “hEx3” in Patent Document 1) showed extremely strong anti-tumor effects. It was further speculated that the structural stability of the variable regions of the above antibodies 528 and OKT3 themselves and their combination are very important for showing such advantageous effects by comparison with an diabody-type bispecific antibody prepared using other antibodies.
Furthermore, the present inventors have developed a highly functional bispecific antibody based on said humanized diabody-type bispecific antibody (Patent Document 2).
Methods for the production of bispecific antibodies other than the diabody-type bispecific antibody are described in Non-Patent Documents 3 and 4.    Patent Document 1: Japanese Patent No. 3803790    Patent Document 2: WO 2007/108152 A1    Non-Patent Document 1: Rosenburg and Moore (Ed.), “The Pharmacology of Monoclonal Antibodies”, Vol. 113, Springer-Verlag, New York, pp. 269-315 (1994)    Non-Patent Document 2: Hollinger, et al., Proc. Natl. Acad. Sci. USA 90, 6444-6448, 1993    Non-Patent Document 3: Alt, M., et. al. Novel tetravalent and bispecific IgG-like antibody molecules combining single-chain diabodies with the immunoglobulin gammal Fc or CH3 region. FEBS Lett., 454, 90-4. (1999)    Non-Patent Document 4: Lu, D., et. al. A fully human recombinant IgG-like bispecific antibody to both the epidermal growth factor receptor and the insulin-like growth factor receptor for enhanced antitumor activity. J Biol Chem., 280, 19665-72. (2005)